24 April 2000
NPT Conference, New York


Nuclear Weapons, NATO, and the EU

von Otfried Nassauer



Nuclear sharing is one the hidden controversial topics that comes up at this NPT-Review Conference as an important point of dispute.Nuclear sharing, which is the agreement that non-nuclear NATO members might use nuclear weapons during times of war, was already controversial when the treaty was negotiated.Historical evidence indicates that, at the time of negotiating and signing the treaty, many countries didn’t fully understand what implications nuclear sharing had and/or didn’t know that NATO interpreted nuclear sharing to be legal under the NPT.According to the current understanding of most Non-NATO states, parties to the NPT, NATO nuclear sharing probably violates Articles I and II of the Treaty.

Let me first read to you these two articles:

Article I says, “Each nuclear weapons State Party to the Treaty undertakes not to transfer to any recipient whatsoever nuclear weapons or other nuclear explosive devices or control over such weapons or explosive devices directly, or indirectly; and not in any way to assist, encourage, or induce any non-nuclear-weapons State to manufacture or otherwise acquire nuclear weapons or other nuclear explosive devices, or control over such weapons or explosive devices.”

Article II, which mirrors that commitment from the non-nuclear weapons sides, states:

“Each non-nuclear-weapon State Party to the Treaty undertakes not to receive the transfer from any transferor whatsoever of nuclear weapons or other nuclear explosive devices or of the control over such weapons or explosive devices directly, or indirectly; nor to manufacture or otherwise acquire nuclear weapons or other explosive devices; and not to seek or receive any assistance in the manufacture of nuclear weapons or other nuclear explosive devices.”

This is complicated treaty language. I hope that I can bring some light into what they might mean, how they have been interpreted, and what the problem around nuclear-sharing is. I’m going to separate my topic into four sections. First, I would like to clarify what NATO nuclear-sharing means. Second, I will talk about the NPT and nuclear-sharing – how it was discussed during the 1960s when the treaty was negotiated. And third, I will address the European Union and nuclear-sharing, which is one question that might come up in the future. Finally, I will conclude with how to circumvent the problem, if one wants to circumvent it.

In their working paper for this NPT Review Conference, the non-aligned countries have made one comment on Article II. Only one. I would like to read it to you as well:

“Non-nuclear-weapons States Parties to the NPT reaffirmed their commitments to the fullest implementation of this Article and to refrain from nuclear sharing for military purposes under any kind of security arrangements with nuclear weapons states, non-nuclear weapons states and states not party to the treaty.”

The latter refers to Israel, the former to the NATO nuclear sharing model. If the majority of the States Party to this Treaty make such a more or less sharp and clear command, then it is absolutely evident that nuclear sharing is a concern for them, whether it is in violation of either the spirit or the letter of the treaty. First, let me explain nuclear sharing.

Six non-nuclear NATO countries are hosting US nuclear weapons on their territories – up to 180 B-61 Model 10 freefall bombs. Some of these bombs are to be used in wartime by the non-nuclear weapons states. Their air forces operate so-called dual-capable aircraft. Dual-capable aircraft have some specific technical design features that allow them to drop not only conventional but also nuclear bombs and to be included in the Command and Control structures for nuclear weapons. The pilots for these aircraft are provided by the non-nuclear weapons states. These pilots undergo specific training, which means that they are trained to use nuclear weapons. Their wings, the air force units to which these pilots and aircraft belong, have the capability to play their part in NATO nuclear planning, including assigning a target, selecting the yield of the warhead for the target, specifically planning a mission, etc.

So, under NATO nuclear sharing, the US would hand over control over these nuclear weapons in times of war to the non-nuclear weapon states’ pilots and for use with aircraft from non-nuclear weapons states. Once the bomb is loaded aboard, the right code for the Permissive Action Link code has been entered by the US soldiers guarding the weapons and the aircraft begins its mission, control over the respective weapon(s) has been transferred. That is the operational, technical part of what is called nuclear-sharing . There is another element to nuclear sharing that I would also like to explain. I would call it the political part of nuclear sharing, which is that all non-nuclear weapons states members to the NATO treaty are eligible to participate in NATO’s nuclear-planning and consultation processes, which means they are eligible to participate in drawing up target plans, in discussing the use of nuclear weapons in war time, in consultations whether NATO should ask the US for the use of nuclear weapons, and in consultations when the US decides to use nuclear weapons, whether NATO would agree to do so. All of these tasks are accomplished in NATO’s Nuclear Planning Group (NPG) and its subsidiary bodies.

NATO nuclear sharing as far as the technical part is concerned was described in 1964 by one member of the National Security Council in the United States, in what was at that time a highly classified memorandum, as meaning that “in effect the non-nuclear weapons become nuclear powers in times of war”. The concern is that the control of the weapon is turned over from the US, as a nuclear weapon state, to non-nuclear weapon states in Europe in times of war, the moment the aircraft loaded with the bomb is on the runway ready to start. The control over this weapon in the physical sense, as well as in the legal sense, is with the pilot from the non-nuclear weapons state. To my understanding, this is in violation of the spirit if not the text of Articles I and II of the NPT.

This is exactly one of those things that the authors of the treaty tried to avoid. Let me go into the history to make it clear why this is a serious problem. During the negotiations for the NPT, NATO’s member states were very careful that this treaty would not prohibit their established system of jointly deciding and implementing specific aspects of NATO’s nuclear strategy and jointly working on nuclear issues. They did so by using a more or less tricky structure. Once the text of Articles I and II was known, the US (in coordination with its allies) worked on a unilateral interpretation of Articles I and II, which they agreed upon internally and then consulted with some of the other countries negotiating the NPT. Who was consulted was widely unknown until recently. We now know that the Russians had been shown the text of these interpretations and that some key members of the Eighteen Nation Committee, negotiating the NPT had been consulted. However, it is still not known, which nations were among the “key” members. Most of the States that signed the NPT on July 1, 1968 had not had a chance to see the text of these interpretations until – at the earliest - nine days later when it was used in public hearings of the US Senate considering the NPT for ratification.

These interpretations were contained in a letter from then-US Foreign Secretary Rusk, who described answers to ‘hypothetical’ questions asked by the European NATO allies. In this letter, he answers four questions. The first three of them are dealing with nuclear-sharing. The fourth one with the future of the European Union. This question-and-answer document, which I will call the Rusk Letter, tries to make legal under the NPT what NATO did anyway. The Alliance did not try to implement or introduce any new elements. They only tried to legitimize under the treaty what NATO already did. The letter argues from the basis that the NPT-treaty does not talk about what is allowed, but only what is forbidden. Everything that is not forbidden by the treaty is perceived by the interpretation to be allowed. This was the general assumption on which this letter was based. Since the treaty doesn’t explicitly forbid the US or other nuclear-weapon states to sell nuclear-weapons-capable carrier systems, such as aircraft, missiles, etc., to non-nuclear weapons states, it is allowed to sell them. Since the treaty doesn’t explicitly talk about the deployment of nuclear warheads in countries that are non-nuclear weapons states, such deployments are considered legal under the NPT. And since the treaty doesn’t talk explicitly about whether it applies or is binding in times of war, a very specific argument has been developed so NATO can argue that this treaty is not binding in times of war. I’ll come back to this aspect a little later.

First, I need to explain one more step - where does the transfer over the control of nuclear weapons take place? The weapons deployed in Europe are guarded by US soldiers and only the US soldiers can activate them by entering a specific code into the Permissive Action Link, PAL, which is a major safety feature in nuclear weapons that is intended to prohibit unauthorized use. If an incorrect code is entered on several tries the weapon renders useless. No ally is in the possession of these codes. Thus the weapons are completely under US command and control up to the point when the US President decides that these weapons can be used and thus the correct codes haven been entered, i.e. the safety switches have been turned the other way around and the weapons are loaded onto aircraft. There’s no major problem with the NPT in this area. The only point where the problem arises is the moment in which a weapon is actually transferred to the non-nuclear weapons state – that is, when the aircraft is going to take off. Beforehand, and before the decision by the U.S. President is taken, etc., there is no clear contradiction to the NPT. >But in that moment, there would be one.

So the question of whether the treaty applies in times of war or not is a very, very crucial one. At the time of negotiating the treaty, the United States and its NATO allies decided not to make these unilateral interpretations available to the other signatories in the normal manner. The normal manner would have been to publish these interpretations as a legal reservation jointly with the signature of the treaty and deposit them with the depository countries. Instead, they were published several days later as materials for the US Senate. The then-leading US diplomat Adrian Fisher argued in the Senate hearing that the US now assume that these documents are known to everybody, to all signatories of the treaty. Vice versa this implies that most of the signatories of July 1, 1968 were unaware of the interpretations contained in the Rusk letter, when they signed the treaty. In addition one has to assume, that surely not all of them were present at this US hearing. Thus many, if not most signatories may have gained knowledge about the unilateral interpretation contained in the Rusk letter until after the hearings were printed, quite a while later. However from the US and NATO point of view this procedure assured that no early questions about NATO nuclear sharing could be raised, simply because other countries didn’t really and fully understand what NATO nuclear sharing meant in detail.

The Americans had clearly agreed to this interpretation, at the time. Journal letters, which have since been de-classified, describe to some of their officials traveling around the world what happened. These documents make it clear that they were asked not to explain the US understanding in too much in detail because objections from the other countries shouldn’t be made easier by giving them too much knowledge about NATO’s nuclear sharing system. One such document states directly that the Soviets did not object since they didn’t have sufficient knowledge on what nuclear sharing technically meant for NATO – even though they had been shown the text of the Rusk letter in advance. They simply couldn’t object to what they didn’t know in detail.From the declassified documents is obvious that something tricky and clever happened here.

Let me now come to the question of whether the treaty is binding in times of war or not. Adrian Fisher, the diplomat who developed the idea for this part of the negotiations, again was acting very clever. He suggested to refer in the preamble of the treaty to the fact that this treaty not only prohibits proliferation, but also prohibits war. Fisher’s suggestion went on to argue, that if such a formula was contained in the preamble it would allow for an interpretation, that once a war had begun, the US could argue, that the treaty had failed to fulfill its function of prohibiting a war and thus was no longer binding to the US and its allies.

The suggestion was adopted and is now contained in the treaty text, which declares that the treaty is intended to prohibit “such a war”. “Such a war” meaning nuclear war. The US interpretation again can be found in the Rusk letter.

In its unilateral interpretations contained in the Rusk letter, the United States explained explicitly, together with its allies, that this means that NATO doesn’t need to feel bound to the treaty in times of war because one major purpose of the treaty is to avoid such a war and this purpose can no longer be fulfilled in times of war.Since NATO has a first-use policy, this means more or less automatically that NATO reserves the right to define when such a war can no longer be avoided.

In the Senate hearing, Senators became suspicious about the trickiness of this argument and asked the question, ‘What type of war are you talking about?’ The Administration replied that they were talking about general war. However, general war is defined in US military strategy; to the opposite NATO’s dictionary doesn’t use or define the term. One of the Senators followed up with: ‘but does that mean every war?’ It was explained, no, it doesn’t mean any war because we assume the treaty would remain in force in times when two minor powers are fighting a war. But when it comes to an East-West conflict, we think that’s general war and in general war NATO wouldn’t be bound to the treaty. I already made clear why the first-use policy makes a big difference here, basically allowing NATO the flexibility to decide, when the NPT should apply and when not.

To sum up these arguments on the history of negotiating and interpreting the NPT with the help of the Rusk letter: This was at least a very tricky and clever way to proceed. It is highly likely that the non-NATO signatories of the Treaty did not have a clear and full understanding of NATO’s reading of the treaty. And it is likely that at least some of them would have raised substantial objections if they would have had a full and clear understanding. I’m not alone on questioning the way the US and NATO proceeded with their interpretations. The then-legal advisor to the State Department, Leonard Meeker in one of his internal memorandums warned the US delegation about negotiating this way. He wrote: “Should we decide to leave the wartime exception implicit we would want to make perfectly clear at Geneva what we are doing, lest we later be accused of having negotiated a treaty under false pretenses”. However, his concerns were obviously neglected. NATO has since had a nuclear-sharing policy, which to my understanding violates the spirit and probably also the letter of the treaty.

Recent developments in NATO make things even more complicated. NATO is currently working on a new military strategy document called MC-400/2, in which the Alliance retains the option to assign to nuclear weapons a role in deterring biological and chemical weapons owners, plus the owners of the delivery means for such weapons. The document is up for political approval in the very near future. Since it is classified, we do not know the exact language. However, as we know from the example of NATO’s first use policy, in NATO an option to use nuclear weapons is perceived to be retained unless the Alliance explicitly excludes it. Whoever assigns to nuclear weapons a role against biological and chemical weapons, not only re-legitimizes nuclear weapons, but also takes care that a first-use policy is a indispensable prerequisite for being capable to implement such a policy. Retaining the option to use nuclear weapons against opponents armed with biological and/or chemical weapons increases the number of occasions under which NATO might consider nuclear-sharing, might consider that non-nuclear weapons states participate in nuclear missions. This is a logical consequence of the Alliance’s policy of shared risks, roles and responsibilities.

In addition: Since we are no longer only talking about the East-West conflict, which is fortunately gone, the wartime exception clause, contained in the unilateral interpretations of the Rusk letter, thus might be deliberately applied to additional situations. This would result in a whole range of new problems, consisting of both possible future violations of Articles I and II and possible violations by the nuclear weapons states of the negative security assurances issued to the non-nuclear weapons states. In their Negative Security Assurances, the nuclear weapon states did not insist on an exception for owners of biological and/or chemical weapons and thus they would violate these assurances in case they would use nuclear weapons against owners of such weapons who do not own nuclear weapons.

Let me finally look into the future and discuss why nuclear sharing might become a problem for the European Union, too. The European Union is likely to face questions similar to NATO at some time in the future - as soon as it starts discussing collective defense. The Union’s members will have to decide whether they integrate their military forces into a collective defense structure or even whether they are going to become a unified state with unified armed forces. In both cases the question will come up, how the members of the Union or a future united state will decide on the use of nuclear weapons. Two of the European Union members are nuclear powers, Britain and France. Thus the “nuclear issue” as it is often called, cannot be avoided.

European integration is often running down a slippery slope. On controversial and possibly tricky topics, such as the “nuclear issue” there is a high likelihood of such a procedure, since a one-time decision to hand-over the control from the national, i.e. British and French, control to the European level is very unlikely. Interim steps, e.g. some version of nuclear sharing modeled after NATO, could be used to avoid a clear-cut decision on a highly complicated issue such as the future control over British and French nuclear weapons. The policy options available to the European nations to make up their mind on this issue will have a strong impact on whether the European Union will finally become a non-nuclear member to the NPT or a nuclear member to the NPT. At least one should try to ensure that the European Union doesn’t run the risk of causing suspicions about the Union violating the NPT in a manner similar to NATO.

Here are my conclusions that we could draw from all this.

First, this review conference should look at the question of whether all members of the treaty can agree to a joint interpretation of NATO nuclear-sharing’s compliance with the NPT. This is the minimum, the absolute minimum. NATO countries should agree that they will adhere to this interpretation.

Second, NATO’s non-nuclear countries should consider whether they will take the unilateral initiative to give up the technical capability to use nuclear weapons. This could be a very, very positive step for strengthening the NPT because it eliminates the ambiguity on whether these countries are in compliance with Article II.

Third, both the non-nuclear as well as the nuclear State Parties to the NPT should consider strengthening and reiterating one formula that some prudent Swedish delegate brought into the 1985 third Review Conference final document, which is that this treaty is binding under any circumstances, including times of war. Then part of the ambiguity created by the US and its NATO allies is gone as well.

My fourth conclusion is that the non-nuclear and the nuclear members of the European Union should assure the other members of the NPT that the European Union is not going to develop at any time a nuclear-sharing model that might violate Articles I and II and thus make clear its very strong commitment to strengthening the non-proliferation regime.


Questions and Answers: Article I & II of the NPT:

Question: The question is about the non-nuclear host countries in NATO.  Is there any objection from people in those countries to having American bombs stationed in their countries.  It seems to me it’s almost a Greenham Common kind of situation.  It has analogies to that, because they become targets when they have nuclear weapons on their soil.  Is there any citizen protest or action that goes on in these countries?

Otfried Nassauer:  Surely there is, even if on small scale.  But the fact is that these countries knowingly become targets.  NATO openly says that shared risks, roles and responsibilities is what makes up NATO’s nuclear strategy.  Sharing the risk means becoming a target and sharing the roles and responsibility means sharing in the use of nuclear weapons.  I think that it is clearly intentional and part of what NATO calls the trans-Atlantic solidarity.


Question:  Question to Otfried.  There are new NATO members that joined last year.  Are there any commitments from the NATO side to not extend nuclear policies to these new members?

Otfried Nassauer:  Yes.  They are politically, but not legally, binding.  First, NATO issued several statements in the NATO-Russia Founding Act and elsewhere to not have any plans, intentions or reasons to deploy nuclear weapons in those new member countries or to use or build infrastructure for nuclear weapons in these countries. Martin [Butcher] and I had a Senator raising the question in the Senate whether this meant that there are no plans or intentions, etc. to conduct the other elements of nuclear sharing - like training fighter pilots and concluding  programs of cooperation, which are legally binding agreements between the US and the individual NATO member and are a prerequisite for sharing nuclear weapons related information.  Both the Secretary of Defense and the Secretary of State in the US clearly answered that the same three “no”s applied to these other elements of nuclear sharing.  So the new members are indeed somewhat second class members to NATO because it would take some time until the politically binding situation as it is right now could be changed.  You need some time to develop and sign and ratify a program of cooperation or make it legally binding between the two countries.  So there is, let’s say, a safety cap.  On the other hand, NATO’s new members are fully eligible to participate in NATO’s Nuclear Planning Group and thus in the Alliance’s nuclear consultations.

 In order to have all NATO members in one class, it would be better if those NATO members, currently having the technical capability to use nuclear weapons, would give that up deliberately and say “in a post-Cold War world, we don’t need that.”  This would ease things very much.  First, all non-nuclear NATO members would be in the same class.   There would be no separation between them any longer; they all would be eligible to join discussions in the NPG, but there would be no longer non-nuclear countries that had the technical capability to use nuclear weapons.  Second, it would strengthen the NPT and non-proliferation regime in general and, to my understanding, third, it would also help Arjun’s proposal - which is making it easier for some foreign ministers to write a letter to the US President saying, “Oh, by the way, on behalf of our country, there’s no need for using nuclear weapons first, ever.”



ist freier Journalist und leitet das Berliner Informationszentrum für Transatlantische Sicherheit (BITS).